The overarching goal of this program project grant application is to change current clinical paradigms through the support of accurate, early detection and measurement of breast cancer response to chemotherapy (Projects 1, 3), and presurgical motor and temporal language cortex evaluation of brain function (Project 2) through the development, application and testing of robust and sophisticated registration and related signal processing tools. Project 1 investigates the efficacy of early-assessment and measurement of response to neoadjuvant chemo or hormonal therapy for patients with breast cancer obtained through the use of volumetric, diffusion and dynamic contrast enhancement MRI. The hypothesis is that nonlinear registration of interval breast exams increases the sensitivity and specificity of functional diffusion mapping (fDM) as well as the accuracy of dynamic contrast enhancement (DCE). Developing low noise, unbiased tools for assessing lesion response to therapy is currently an important topic. Project 2 extends previously completed work on registration-based fMRI motion by examining the benefits of combining our unique motion correction method with different fMRI acquisition protocols, e.g. clustered acquisition, to improve communication with the patient and response monitoring. Project 3 addresses the fundamental ambiguity problem in dynamic MRI associated with imaging in general: for any given technique either we can obtain high spatial or temporal resolution imaging data, but not both. Generalized techniques that support controlling and optimizing these tradeoffs during dynamic imaging in MRI are very important.